Case Study: AstraZeneca predicts pharmacokinetics with Intellegens Alchemite

Predicting pharmacokinetics with Optibrium and AstraZeneca

AstraZeneca partnered with Intellegens to improve prediction of pharmacokinetics (PK) in drug discovery. The challenge was to better estimate how compounds behave in the body, including concentration-time profiles, so the team could make stronger decisions about which compounds to move into in vivo studies while reducing cost, time, and animal testing. Intellegens used its Alchemite™ machine learning approach to work from chemical structure and sparse in vitro data.

Intellegens’ solution successfully combined descriptors, in silico data, and in vitro data to predict PK parameters and time-exposure curves. The results were comparable to the best published methods and AstraZeneca’s in-house approaches, with direct prediction of i.v. curves being especially accurate. This has the potential to reduce late-stage drug discovery effort, costs, and the number of animal studies required.

AstraZeneca

Nigel Greene

Director of Data Science & AI